“our ai broke out and hacked some organisations” = "we failed to set up appropriate safeguards, failed to continuously monitor our experiments, and ignored a basic principle of reinforcement learning: if hacking gets rewarded, the system learns to hack"
216 reposts773 likes